N.Solyak, RTML Introduction ILC LET workshop, CERN, June 23-25, 2009 1
ILC RTML IntroductionNikolay Solyak
Fermilab
RTML function and RDR design
Minimum machine and Single-stage BC
Emittance preservation in RTML
Stray magnetic fields studies
Slow ground motion Studies at FNAL
Summary
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 2
Note: e- and e+ RTMLs have minor differences in Return line (undulator in e- linac side), they are otherwise identical.
BDS
RTML/Sourcetunnel
Areas, where tunnel length saving is possible
Same curved tunnel ~12 km (RTML/ML)
1254 m
1208 m
RTML Schematic (RDR)
RTML Laser Straight tunnel
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 3
RTML Functions
• Transport Beam from DR to ML – Match Geometry/Optics
• Collimate Halo
• Rotate Spin
• Compress Bunch
• Preserve Emittance
– vertical norm. emittance < 4nm
• Protect Machine
– 3 Tune-up / MPS abort dumps
• Additional constraint:– Share the tunnel with e-,e+injectors – Need to keep geometries
synchronized
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 4
RTML Optics Design (RDR)
DRX connection
Vertical plane
Horizontal plane
Horizontal plane
• Horizontal Arc out of DR ~1.1 km straight
– In injector tunnel• “Escalator” ~0.6 km vertical dogleg
down to linac tunnel• Return line (weak FODO lattice) ~13km
– In linac tunnel– Vertically curved
• Vertical and horizontal doglegs• Turnaround• 8° arc in spin rotators• BCs are net straight• ML launch
DR-RTML hand-off point defined extraction point where η,η’ 0RTML mostly defined by need to follow LTR geometry Stay in same tunnelDesign is OK at conceptual level
DRX+ arc
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 5
DR connection (RDR)
•Both sides have CMs for sources
• e+ side also have KAS and e+ transfer line from undulator
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 6
DRX Connection
• Current design is entirely planar (horizontal plane)
• DRs are in different planes
• Sources need cryomodules and SC solenoids– Big heavy objects which
want to sit on the floor
• Working agreement between sources, DR, RTML, CFS:– CMs and SC solenoids
always sit on floor– RTML hangs from source
tunnel ceiling at same location as in linac tunnel
DR Tunnel – 1.44 m Vertical separation
ML Tunnel - 2.14 m Vertical beam
separation
e+
e-
DR tunnel
e-
e+
e- src
e+ RTMLe- RTML
e+ src
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 7
“Getaway” Straight (or “DR Stretch”)
• About 1.1 km long• Has two parts
– “Low-beta” region with decoupling and emittance measurement
– “High-beta” region with collimation system
• Includes PPS stoppers– For segmentation
• Good conceptual design– Need to match exact
required system lengths– Need to consider
conflicts with source beamlines in this area
– Beta match between low- and high-beta optics not great
Beam collimationEnergy collimation
Diagnostics: Emittance meas
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 8
Escalator
• Vertical dogleg–Descends 7.85 meters over ~590 m
–Uses 2 vertical arcs separated by weak FODO lattice
• Good conceptual design–Escalator-linac tunnel connection does not match CFS design
–Uses Keil-style eta matching
–Beta match between “strong” and “weak” lattices not great
–Positron return line confilicts?
• Need to make match according CFS (new?) design
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 9
Return Line• Weak FODO lattice at ML
ceiling elevation (1Q/~36m)• Vertically curved tunnel thru
ML area– Dispersion matching via
dipole correctors
• Laser-straight tunnel thru BC area
• Electron line 1.2 km longer than positron– Goes thru undulator area
• System lengths probably not exactly right
• Electron Return line and positron transfer line need to be exchanged
e- ML
Undulator 400 MeV e+
e- Return
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 11
Spin Rotation• Design based on Emma’s
from NLC ZDR– 2 solenoids with Emma
rotator between them• Rotate spin 90° in xy plane
while cancelling coupling
– 8° arc• Rotate spin 90° in xz plane
– Another 2 solenoids + Emma rotator
• Basic design seems sound– Very small loss in
polarization from vertical bending in linac tunnel
• Important issue = bandwidth– Off-energy particles don’t
get perfect cancellation of dispersion and coupling
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 12
ILC Baseline 2-stage Bunch Compressor
• Longitudinal emittance out of DR:– 6mm (or 9 mm) RMS length– 0.15% RMS energy spread
• Want to go down to 0.2-0.3 mm• Need some adjustability• Use 2-stage BC to limit max energy
spread– 1st: Compress to 1 mm at 5 GeV– 2nd: Accelerate to 15 GeV and Compress to final bunch length
• Both stages use 6-cell lattice with quads and bends to achieve momentum compaction (wiggler)– Magnet aperture ~ 40cm
• Total Length ~1100 m (incl. matching and beam extraction lines)
• Minimum design is possible if assume compression 60.3 mm only
• Shorter 2-stage BC• Or short single-stage BC• Cheaper magnets
RF system• BC1: 3 CMs with quads/each (+spare kly)• BC2: 14 RFunits (3CM’s each)+1spare • Total 48 CM’s per side
One wiggler cell
RF1 RF2Wiggler 1
Wiggler 2
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 13
Cost and its Distribution
• CFS + BC RF system = 68% of costs– Correlated – much of CFS
cost is housing for BC cryomodules
• Remainder dominated by RT beam transport– Quads, correctors, BPMs,
vacuum system
• Small amount of “exotica”– Non-BPM instrumentation,
controls, dumps, collimators
CFS
CM
RF
Cryo
Instrumentation
Dumps + Colls
Vacuum
Magnets + PS
Controls
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 14
RTML Risk Analysis (2007)
(1) Packing fraction to tight (Turn-around) Q* Med 12
(2) Beam Motion at end of RTML Too Large Low 2
(3) Optical Effects in RTML Low 5
(4) Space charge tune shift in RTML Med 5
(5) Ion Instability in e- line Low 5
(6) Collimator Wakefields Low 5
(7) Emittance Dilution : Magnetostatics High 10
(8) Emittance Dilution Pitched RF Cavities High 20
(9) Bunch Compressor Phase Stability Med 60(10) Performance of Feed-forward Med ?
Risk Cost
*Q - CONCERNS REQUIRING SIGNIFICANT CF&S MODIFICATIONS AND DECISIONS EITHER DURING EDR OR SHORTLY THEREAFTER
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 15
Minimum Machine configuration
• Cost-reduction strategy (2009)• Major changes vs. Baseline configuration
– ML• removal of the underground service tunnel (single underground
tunnel housing the accelerator);• klystron cluster concept (RF power distribution alternative);• processed water cooling specifications (higher ΔT solutions).
– DR• Shorter DR for Low power option (1320 bunches vs. 1625)
– RTML• Single stage BC, Remove one Extraction Line with dump
– Source and BDS Integration (Central Region)• Move undulator to 250 GeV• 10% KAS• Sources, undulator and 5 GeV e+/e- Linacs in BDS tunnel• Shorter BDS
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 16
RTML changes in Minimum Machine Configuration
The RTML two-stage Bunch Compressor (top) and a possible short single-stage compressor (bottom). Lengths compared
for 15 GeV.
~200-250 mG=31 MV/mPhase = -5 deg
14+1 RF Units (45CMs)
14 RF Units (42 CMs)4.6 GeV Main Linac
Single-stage BC is possible, if not support flexibility of parameter set
Changes from RDR: 9(6)mm0.3(0.2)mm to 6mm 0.3mm (x20 compression) Reduction in beamline and associated tunnel length by an equivalent of
~200-250 m (including some in SCRF linac) Removal of the second 220 kW dump and dump line components Shortening of the diagnostics sections (lower energy)
12 RF (36 CMs)2 RF
1 RF
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 17
Single-Stage BC Lattice
Bunch Length: - 6 mm → 0.3 mm
Total length = 423 m- RF Section- Wiggler- Diagnostics & EL- Matching section
Final energy 4.3 GeV
Pre-Linac: 4.3 GeV 15 GeV
Based on the original design, proposed by PT et al. in April 2005:
PT lattice is modified to improve performances:
• CM-3 CM-4
• Wiggler wiggler 2007 PT/Seletskyi
(see: A.Latina talk)
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 18
Design Characteristics
423 m
E
N
3.54 %@ 4.3 GeV
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 19
•ELBC2 length ~25 m (longest one)•6 septum+6 bends+12 quads, •two collimators: 5.2 kW (protect quads) and 14.1 kW (dump window)
•10 fast kickers and pulsed bend in the main beamline to extract beam
•Beam dump 220 kW @ 15 GeV0.15% (green) and 1.8% (red) energy spread
2 coll 1 coll No coll
Final quads
1T 45mm
1T 45mm
2T80mm
Collimat 5.2 kW 14.1kW
5.2kW
No coll
Dump window
12.5 cm 30 cm
100 cm
Defocus
Removing of ELBC2 and dump
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 20
~220kW Aluminum Ball Dumps
Cost (~$1M each) is dominated by:– 3-loop radioactive water
processing system– The CFS infrastructure,
shielding, etc.Similar dumps in use at SLAC
50cm Diameter x 2m long Aluminum Ball Dump with
Local Shielding
RW
50kW 3-loop 2006 Rad Water Cooling for ISIS Neutron Spallation Targets
Remove 2 Dumps after BC2
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 21
Re-design of the ELBC1
• Motivation: Accommodation of larger energy spread (3.6 % vs. 2.5% in previous design)– For the beam with high energy spread, there is a substantial blowup in
the beam size from chromaticity and nonlinear dispersion at the end of the beamline.
• Few options were studied (TILC09, S.Seletskiy)– No collimation, sextupoles at the beginning/end – No collimation, sextupoles at the end– Weak collimation and Sextupole– Strong collimation with 2 collimators
• Needs more studies with experts to choose final design. Decision for the final design must be taken through cost optimization process.
See more details in S.Seletskyi talk
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 22
ILC New Damping Ring• New ILC DR lattice is shorter.
• Bunch length = 6 mm; Energy spread = 0.13 %In old RDR design: • 9 mm (easy)• 6 mm (more challenge)
• New DR increases the length of the RTML linac in each side (e+ and e-) of ~300 m, but not CFS
• Need redesign/adjust DRX lattice to accommodate changes in DR
• Compact DR (S.Guiducci, DESY AD&I)
- SB 2009 (DSB) – 3.2 km (half- number of bunches)
Injection
ExtractionExtraction
Injection
300 m
Layout of the ILC Damping Ringblue - RDR (2007); red - new DCO (Feb.2008)
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 23
Estimated Cost impact
• CFS: reduction of tunnels:– 210 m of regular tunnel
– No service tunnel
– No alcoves for 2 extraction lines and dumps (radiation area)
– Possible more saving in tunnel in central area
• Cost reduction due to reduction of hardware components (~30-40 M$)– 12 CMs
– 8 klystrons/modulator/PDS
– 2 extraction lines with 2 beam dumps
– Magnets, fast kickers, septums, PS
– Diagnostics: LOLA cavities, BPMs, etc..
– Vacuum components
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 24
RTML in Central Area
• New configuration of the Central Area and DR design will require changes in RTML design– Need to complete configuration ASAP. It will be
basis for RTML lattice design work and cost estimation
• Expected incremental cost due to this changes will be small
• Biggest impact on CFS (tunnel)
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 25
Emittance Growth in RTML
Region BBA method Dispersive or chromatic mean emittance growth
Coupling mean emittance growth
Return Line KM and FF to remove beam
jitter
0.15 nm 2 nm (with correction)
Turn around Spin rotator
KM and Skew coupling
correction
1.52 nm (mostly chromatic)
0.4 nm (after correction)
Bunch Compressor
KM or DFS and Dispersion bumps
>5 nm (KM+bumps)2.7 nm (DFS+bumps)
0.6 nm (w/o correction)
Total ~5 nm almost all from BC
3nm (w/o complete correction)
• Effect of coupler RF kick & wakes is not included• Dynamic effects are not included• Emittance growth is large (pre-RDR budget 4nm, might be
≤10nm)• Need further studies to reach goal for emittance growth• Cross-checking with different codes (important)
Summary of Studies before MM (LET meeting, Dec.2007 SLAC)
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 26
Simulations of Coupler Kick and Wakes
Wx(s)-solid,,Wy(s)-dashed for σz = 300 μm.
On-axis kick factor vs. σz
bunch shape
The couplers break the RF field symmetry and cause transverse RF kick and WakesDESY,2007. Simulations DESY/FNAL/SLAC
The profiles of the 3 couplers, as seen from the downstream end.
Upstream HOM coupler
Downstream HOM coupler
Main coupler
RF cavity
~20%
Total RF KICK
FNAL Q=3.5106
HFSS
DESY Q=2.5106
MAFIA
SLAC Q=3.5106
OMEGA3P
106 · (Vx/Vz) -105.3+69.8i -82.1+58.1i -88.3-60.2i*
106 · (Vy/Vz) -7.3+11.1i -9.2+1.8i -4.6+5.6i
Coupler Transverse Wakefield
Effect of couplers on emittance growth see in V.Yakovlev talk
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 27
Coupler and Misalignments in BC1S
•BC1S (incl. diagnistics+matching+Pre-linac (515 GeV)) •Standard misalignments (300 um/300urad); ISR +coupler RF
kick/wake •1-to-1, DFS and bumps, girder optimization
10 nm 5 nm2.6 nm
2.2 nmfrom
coupleronly
Girder Pitch optimization
Y- micromover: - Range 300 um - Step size 10 um
#N of adjustable CM’s
• RF section of BC1S - 1 every 2 (total 3)
• Pre-linac: 1 every 12 CM’s (total 3)
New proposal !!!
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 28
Emittance Growth in Bunch Compressor
• Emittance growth due to misalignments and couplers seems to compensated both for BS1S and BC1+BC2
• Girder pitch optimization is very effective to counteract coupler kicks, both for BS1S and BC1+BC2
• In BC1S, Crab Cavity seems to be similar effective, but it would require a new hardware and slight redesign of the cryomoodule
A.Latina, TILC09
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 29
Stray magnetic fields (2)
RTML requirement for high frequency stray magnetic fields: B < 2 nT (K.Kubo, 2006)
Magnetic field examples– Commercial SC solenoid – 10 Tesla (1 e+1)– Earth magnetic field – 50 micro-Tesla (5 e-5)– Cell phone – 100 nano-Tesla (1 e-7)– Beating human heart – ~ 10 pico-Tesla (1 e-11)
Frequency dependence– < 0.1 Hz (can be compensated by control system)– > 100 kHz (attenuated in the structure)
Classification (following F.R.T.)– 60 Hz and its harmonics (near-coherent with 5-Hz pulsing)– Fields from RF systems (coherent with 5-Hz pulsing)– Others (non-RF technical sources) (uncorrelated with pulses)
Previous work- “Sensitivity to Nano-Tesla Scale Stray Magnetic Fields”, SLAC LCC Note-0140 (June 7, 2004) Data from SLC End station B.
Conclusion (for NLC): we are mostly OK- DESY (TESLA TDR study)
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 30
Stray Magnetic Field studies 2007 plans
RTML needs to transport low-emittance 5 GeV beam over ~15 km from DR to ML. RDR requirement on stray magnetic fields in the RTML is less than 2 nT.
Proposal summary:
• Evaluate possible sources of the stray fields; consider effects of correlated and uncorrelated (with the beam cycle) sources.
• Survey the existing sites to verify assumptions in that analysis. • If the result of this study would require, propose shielding approach for the beam pipe. • Develop a stray field model suitable for incorporation into linac simulation frameworks.
Implementation plan (2007):
• Anticipated duration of this work is two years. The scope of the work can be subdivided in the following interleaving parts:
1. Evaluate the expected magnetic field from multiple sources in the RTML environment: 1.1 Correlated sources: RF systems. 1.2 Power-line (60 Hz) and its harmonics. 1.3 Miscellaneous random noise
2. Survey multiple sites in existing installations (Fermilab, CERN, SLAC, etc..) to verify assumptions in stray field model. To accurately measure magnetic field in a wide range of frequencies and amplitudes from 10 microtesla (Earth magnetic field) to below nT is a challenge. We will need to develop a system that can perform this task.
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 31
Magnetic Stray Fields Studies
Hardware: •3-axis fluxgate magnetometer• 0.1mT full scale• DC to 3 kHz • 20 pT/sqrt(Hz)
Measurement:• Near klystron• In shielded cave (20m from kly)• Klystron On/Off
Fermilab A0 experimental area with cryogenic and 5 MW klystron/modulator
•RTML requirement for stray fields in Return Line < 2nT (freq>1Hz)•SLAC measurements (at Station A) are promising (~2nT)•Need more studies for different sites. Stability of 60Hz is an issue
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 32
• Total integrated field ~9 nT ( >1Hz )
•Only 60/180/300 Hz peaks are removed
•15, 30 and 45 Hz (Linac & Booster) ~ 5nT
•Other harmonics 120/180 /240Hz)~ 2nT
•All the rest ~ 2 nT
Spectrogram and
Integrated spectrum
• 5MW Klystron• Modulator• Transformer• Cryogenic• Pumps• Power
Supplies• Tevatron
shaft • Etc…
Stray magnetic fields measurements at A0:
A0 - Noisy place
More details in D.Sergatskov talk
Mu-metal shielding will help to reduce fields !
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 33
Vacuum system for RTML Return line
A A B End
~36m
V V V
V RGA
Vacuum breaker or gate valve
By pass
String~144m
Bellow
CA AA CA AA
Ion Pump
Magnet
A A B End
CA AA CA AA
Ion Pump
Magnet
A A B End
V V V
V RGA
CA AA CA AA
Ion Pump
Magnet
A A BCA AA CA AA
Ion Pump
Magnet
LD
Pmax<8.0E-9Torr
Cell~72m
Cell~72m
•Tight vacuum req. in Return Line (<10nTorr)- Passivated SS, ID=35mm, in magnet ID=16mm - 86 curved strings followed by 8 straight strings; - 1 bellow/1 quad magnet- If one string uses vacuum breaker, the next string uses gate valve.
•Final Report with Cost estimation(Xiao Qiong, IHEP/China)
N.Solyak, RTML Introduction ILC LET workshop, CERN, June 23-25, 2009 34
Slow Ground Motion and Vibration Studies at Fermilab
Goal:
Collect data and Study long term stability (different scales: day, month season, years) of the surface and deep tunnel. Analyze sources of ground motion including technical noise.
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 35
HLS systems at FNAL
• MINOS near detector– 4 BINP sensors – 30 meters apart, 100 meters deep– 11 months of data
• North Aurora Mine with 6 sensors– 6 BINP sensors – 30 meters apart, 100 meter deep – 10 months of data with 2 sensor 2 months
of data with 6 sensors• B0 and D0 interaction regions
– 8 BINP sensors each interaction region– On low beta quads and central detector– 5 months of data
• A-F sectors– 204 Balluff sensors one on each SC quad
around the Tevatron– 30 meters apart– 5 months of data All are two pipe systems
BINP sensor (parts)
Fermilab Balluff sensor
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 36
Difference in two sensors150 meters apart
18 months of data
J T Volk Fermilab Dec 2008
Major RainEvent and Flooding in Mine
9/1/2006 3/1/2007 9/1/2007 3/1/2008600
700
800
900
1000
Mic
or
Mete
rs
Date
Difference in two sensors150 meter apart LaFarge MineNorth Aurora Illinois
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 37
Horizontal measurements
0 2 4 6 8 10 12 14 16 18 20 22 240.000
0.001
0.002
0.003
0.004
0.005
0.006
0.007
Frequency (Hz)
Am
plit
ude
Horizontal MINOS no LCW or fans
Vertical measurements
Seismometer measurements MINOS hall
0 2 4 6 8 10 12 14 16 18 20 22 24 26 280.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
Frequency (Hz)
Am
plitu
de V
olts
Vertical MINOS hall no LCW or fans
Vertical MINOS hallNo LCW or fans
A large spike at 10 and 20 Hz is apparent in the vertical it is assumed that this is due to vacuum and LCW systems in the decay tunnel.
Seismic measurements made in the MINOS hall at depth of 100 meters.
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 38
Ground Motion Summary
• There are several HLS system taking data at Fermilab– Aurora mine; MINOS hall; NML hall.
• They are accurate and reliable can run for several years.
• They are useful for determining ground motion and tilt.
• The data are available at; http://dbweb1.fnal.gov:8100/ilc/ILCGroundApp.py/index
• There are natural sources of motion: tides, rain fall, earth quakes both large and small.
• There are cultural sources such as sump pumps.• Plans for new systems in the works. J.T. Volk, Fermilab
N.Solyak, RTML Introduction ILC LET Workshop CERN, June 23-25, 2009 39
Summary
• Single stage Bunch Compressor is designed and studied. Design looks feasible.
• Emittance growth in bunch compressor can be effectively controlled, by using movers to adjust tilt of the cryomodules. (only few CM’s with this features are needed). R&D is required.
• Extraction line is redesigned to accommodate bunch with a larger energy spread after BC
• Magnetic Stray Fields measurements under way. Results are promising (mu-metal shielding is helpful)
• Slow ground motion data are available for analysis of their effect on emittance growth.
Top Related